1
votes

I modified XTK's lesson 13 to load DICOM images from my desktop and do volume rendering.

Although it is working fine, I noticed that the X.volume object is holding nearly 4 times the total size of pixel data in the DICOM images. Is it an expected behavior, and if not then what is the correct method to do this?

1
it may be using double as representation for the pixel type instead of the original short pixel type (or maybe your DICOM dataset is compressed)malat

1 Answers

0
votes

Found out that XTK initializes 4 unique volume objects, which leads to this behavior. However it needs only the X.volume._IJKVolume object for the final rendering, and rest of the volume objects (i.e. X.volume.slices, X.volume.data and X.volume._IJKVolumeN) are no more needed after X.volume._IJKVolume is created.